Judges and Discrimination: Assessing the Theory and Practice of Criminal Sentencing

Judges and Discrimination: Assessing the Theory and Practice of Criminal SentencingThe author(s) shown below used Federal funds provided by the U.S. Department of Justice and prepared the following final report:
Document Title: Judges and Discrimination: Assessing the Theory and Practice of Criminal Sentencing
Author(s): Charles W. Ostrom ; Brian J. Ostrom ; Matthew Kleiman
Document No.: 204024
Award Number: 98-CE-VX-0008
This report has not been published by the U.S. Department of Justice. To provide better customer service, NCJRS has made this Federally- funded grant final report available electronically in addition to traditional paper copies.
Opinions or points of view expressed are those of the author(s) and do not necessarily reflect
the official position or policies of the U.S. Department of Justice.
Judges and Discrimination
Apprwed By: m
Charles W. Ostrom Brian J. Ostrom Matthew Kleiman
Michigan State University National Center for State Courts National Center for State Courts
With the cooperation of the Michigan Sentencing Commission and the Michigan Department of Corrections
This report was developed under a grant from the National Institute of Justice (Grant 98-CE-VX-0008). The opinions andpoints of view in this report are those of the authors and do not necessarily represent the official position or policies of the National Institute or the Michigan Sentencing Commission.
1
The sentencing decision is the symbolic keystone of the c~minal justice system: in it, the conflicts between the goals of equal justice under the law and indvidualzed justice with punishment tailored to the offender are played out, and sociely5s moral principles and highest vaiues-life and libedy-are interpreted and appled.
Research on Sentencing: n e Search for Reform (1983)
The sentencing decision receives considerable attention from both
researchers-who look at the determinants of the sentence choice-and policy
makers-who look at the necessity and possibility of reform. Yet, despite
decades of study, there are large gaps in our understanding of how
"punishments are tailored" across the full range of sentencing options. Any effort
to explain sentencing outcomes means coming to terms with judicial discretion
and how it is put into practice.
Judges are the voice of sentencing, but their freedom of choice is limited
by the statutes and sentencing structures existing in a particular state. Since the
late 1970s, judicial discretion has been constrained by the creation of sentencing
guidelines and other means for structuring the sentencing decision. Some argue
these structures unduly restrict a judge's ability to appropriately weigh the
factors that play a role in sentencing, while others feel that additional measures,
such as mandatory minimum sentencing laws, are needed to constrain judicial
discretion further. But regardless of ones personal views on circumscribing
judicial discretion, a comprehensive assessment of sentencing outcomes in
2
present day America must integrate understanding of the sentencing proclivities
of individual judges with the way discretion is shaped and controlled.
Our interest in sentencing outcomes is focused on disparity, According to
the U.S. Congress, sentencing disparity exists "when defendants with similar
criminal records found guilty of similar criminal conduct receive dissimilar
sentences."' I n practice, disparity refers to differences in sentencing outcomes
that are associated with "extralegal" factors such as race, ethnicity, economic
standing, gender, or the exercise of certain procedural rights (e.g., trial).
Disparity exists whenever extralegal factors influence the sentence once the
legitimate factors related to the offense/offender are taken into account. Not all
sentencing variation is unwarranted; sentences properly reflect differences in the
seriousness of the offense and/or the criminal history of the offender. Assessing
disparity requires distinguishing between differences in sentences that meet the
state's legitimate purposes in sentencing and those differences whose
consideration is prohibited by the offender's constitutional rights.
Existing data sources make clear why the issue of disparity in sentencing
remains center stage. Between 1980 and 2000 the prison population in the
United States grew by one million individuals. As can be seen in Figure 1-1, the
rate per 100,000 population in 1980 was 139, growing to 460 per 100,000 by
2000 - an increase of 300%. Dramatic increases have been experienced in all
states.
figure 1-1 : Incarceration rates for prisoners under State or Federal jurisdiction, per 100,000 residents
e 500 0 3 450 (P g 400 2 350
8 250 200
t o
8 300
a .- J , , , , , , , , , , , , , , , , , , I 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998
Year
Figure 1-2 shows that the increases in prison population have not
occurred across the board. Although violent offenders comprise the largest
share of prison population, the greatest increase has been for drug offenders. In
1980, 58% of those coming to prison were convicted of violent offenses and 6%
for drug offenses. By 2000,48% were coming for violent offenses while 21%
were coming for drug offenses. Proportionally, there are lO0/o fewer violent
offenders and 15O/0 more drug offenders than in 1980.
4
Figure 1-2 Number of Prison Sentences by Offense Type, 1980-1999
600,OOo
500,000
400,OOo
300,OOO
200,000
100,OOo
a I 1980 1983 1986 1989 1992 1995 1998
The racial mix of prison population has remained remarkably stable over
the past two decades. While blacks make up 12S0/o of the US. population, they
account for about one half of the prison population: African Americans are still
over-represented in state prison population versus the total population in all
states.* The average incarceration rate for Blacks is 1,547 per 100,000
population while the average for Whites is 188 per 100,000 -the incarceration
rate for Blacks is approximately 8 times higher than for Whites. On the surface,
these numbers raise the specter of discrimination.
Our empirical approach to assessing disparity is illustrated through a
comprehensive analysis of sentencing outcomes in Michigan. Sentencing in
Michigan is structured through a set of sentencing guidelines that are described
in detail in Chapter 5. While we believe our methods have utility for measuring
disparity in all sentencing systems, the choice to develop and refine them within
* In a recent Bureau of Justice Statistics (BJS) special report (Bonczar et al. 1997) using 1991 as a base year, it was noted that Blacks have a 28.5% chance of going to prison in their lifetime, Hispanics have a 16% chance, and whites have a 4.4% chance.
5
a guidelines context offers several advantages (see also Ulmer 1997). To begin
with, guidelines are an accepted part of the sentencing landscape and viewed by
many as the best means to structure judicial discretion to attain consistency and
fairness in sentencing. I n addition, the control of disparity is often an explicit
rationale for their development. Finally, guidelines require the collection and
classification of a great deal of objective data essential for the empirical study of
disparity. Therefore, beyond clarification of the statistical methods, our
approach allows us to address the larger issue of how the guidelines model of
structuring judicial discretion can be designed to better achieve the articulated
goals of consistency and elimination of unwarranted disparity-with data
sufficient to conduct a full and comprehensive study.
We begin with an overview of the recent history to control judicial
discretion through sentencing guidelines. It also serves to place the Michigan
system in broader perspective. We then turn to the specific goals of this book.
The Changing Nature of Judicial Discretion
Judicial discretion is increasingly constrained through sentencing
guidelines as well as mandatory minimum penalties, three-strikes laws, and other
forms of structured sentencing. The sentencing reforms since the 1970s have
sought to limit judicial discretion and, at least on paper, have largely succeeded
in that objective. Judges in many states find that statutory provisions mandate
the factors that must be considered when passing sentence, designate the
relative importance of those factors, and specify a presumptive sentence (or
6
range) for a defendant based on offense seriousness and prior criminal
involvement.
The attention given to judicial sentencing should be understood as a part
of a sea change that occurred in sentencing philosophies. The 1970s brought
the transition from a venerable system of "indeterminate sentencing" to a new
one organized around the principle of "just desserts." Indeterminate sentencing
combined two main features. First, judicial discretion in sentencing was wide
and largely unchecked, save for legislatively specified maximums and (less
commonly) minimums. Second, judicial decisions regarding sentence length was
paired with a system of state parole boards, appointed by the governor, whose
release decisions determined the actual length of time offenders spent in
custody.
The formal principle underlying indeterminate sentencing was substantive
rationality: achieving the sentence that is just for each individual defendant
(Ulmer and Kramer 1996). This principle encouraged the use of extralegal
factors to establish, for example, the rehabilitative potential of the offender. The
heyday of indeterminate sentencing coincided with a period of optimism about
the potential for rehabilitation--the 1960s--in which treatment displaced
punishment as the official role for penal institutions.
Structured sentencing, including guideline systems, arose in response to
what were perceived as undesirable features of indeterminate sentencing. Some
critics claimed that disparity was promoted when judges were given no guidance
7
in how to incorporate all sentencing-relevant factors in a consistent fashion
(Freed 1992, 1687). I n this regard, it was possible for offenders with identical
offenses and prior criminal records to receive vastly different sentences. Also,
critics noted that indeterminate sentencing did not encourage appellate review of
sentencing decisions (Freed 1992, 1688). Moreover, the broad discretionary
powers of parole boards ensured that judges could never be certain of the actual
time that a convicted offender would serve following imposition of sentence.
With structured sentencing, formal rationality (predictable and uniform
application of rules) replaces substantive rationality as the principle underlying
sentencing. This principle has been most influentially expressed in the just
deserts approach to sentencing (see: Ulmer and Kramer 1996). A just deserts
philosophy was used to establish the first state sentencing guideline system,
enacted by Minnesota in 1980.
Since the mid-l970s, judicial discretion has been out of fashion and parole
boards have been eliminated or have had their discretionary release authority
substantially curtailed. Implementing sentencing guidelines is not the only
mechanism for structuring discretion. Legislatures enact mandatory minimum
sentences in pursuit of uniform and severe punishment of selected types of
offenders (e.g., those committing their crimes with firearms in proximity to a
public school). Habitual offender and "three-stri kes" provisions also direct
judicial sentencing to accomplish specific objectives of the legislative branch.
8
States have pursued various paths to sentencing reform. Many states
have adopted a prescriptive approach to guidelines development, whereby a
sentencing rationale has been articulated at the onset and guidelines have been
formulated consistent with that rationale. Other states have created guidelines
in a more descriptive fashion, such that sentence recommendations more closely
reflect past sentencing behavior of judges in felony cases. However, no state
has gone the route of the U.S. Sentencing Commission, which developed a highly
detailed and mechanical set of guidelines without a clear rationale.
Observers disagree on the overall impact of sentencing guidelines on
judicial discretion. Some feel that the criminal justice system contains a fixed
amount of discretion. Efforts to restrict discretion at any one stage or by any
one category of actor in the system fail because of the "hydraulic displacement
of discretion" (McCoy 1984). That is, a change or reform that limits discretion in
one area leads to an increase of discretion in another area. One view is that
sentencing reform in recent decades has diminished the discretion of judges and
parole offices, but has enhanced discretion of the prosecutor (Boerner 1995). As
the gatekeepers to the system, prosecutors decide which defendants are
prosecuted and with what charges.
More recently, observers have stressed the discretion that judges retain in
the sentencing process. Despite the embrace of just deserts in Minnesota's
sentencing guidelines, "attorneys and trial judges remain firmly attached to
offender-based crime control sentencing goals" and the use of judicial discretion
9
has steadily increased. Some of the prime movers in the sentencing guidelines
movement now claim that enhancing judicial discretion (by reducing "back end"
discretion on the part of parole boards) was one of their original objectives and
note the accountability that judges uniquely possess for their sentencing
decisions (Knapp 1993; Alschuler 1993).
Sentencing Guidelines Basics
Guideline sentences are typically based on factors such as offense
severity, the offender's prior record, the availability of punishment alternatives,
and concerns for community safety. Distinctions exist, though, in how stringent
different guidelines systems are in "limiting" judicial discretion. The limits are
found in the sentencing procedures that direct judges to reference, consider, and
adhere to a specific recommendation on a sentencing grid or worksheet. I n
conjunction with state statutes and authority, these procedures or "mechanics"
define the extent to which a system is voluntary or mandatory in nature.
Sentencing guidelines systems developed during the early 1980s were
quickly categorized by scholars and practitioners as either presumptive or
advisory (mandatory or voluntary) and as either prescriptive or descriptive
(Tonry 1987; Frase 1995; Morris and Tonry 1990). As new systems have
developed, these distinctions have become blurred, with states using
combinations or hybrids of the earlier systems. For this reason, it is difficult to
classify the various sentencing guidelines systems into a rigid
voluntary/mandatory dichotomy. In general terms, most sentencing guidelines
10
systems make use of a grid or set of worksheets that are completed before
sentencing and provided to the sentencing judge. Each guidelines system has
policies and procedures addressing issues such as when guideline forms should
be completed, when judges must review guidelines, how compliance or
departures are to be handled, and what appellate rights are retained by the
defense or prosecution.
Figure 1-3: Policies Governing Sentencing Guidelines’ Use in Selected States
State I Policies
1. North Carolina
State law requires completion of guidelines forms Compliance within presumptive, aggravated, or mitigated ranges is required; no departures
Sentences within approved aggravated and mitigated ranges must be accompanied by
State law requires completion of guidelines forms. Judges must review guidelines forms State law requires departures to be accompanied by written justification Defense and prosecution may appeal sentences
State law requires completion of guidelines forms State law requires departures to be accompanied by written justification Defense and prosecution may appeal sentences
are allowed
written justification
2c. Michigan
2d. Minnemta
~~
State law requires completion of guidelines forms State law requires departures to be accompanied by written justification Defense and prosecution may appeal sentences
State law requires completion of guidelines forms State law requires departures to be accompanied by written justification Defense and prosecution may appeal sentences
11
~
There are no guidelines forms to be Completed Judges are guided by rebuttable presumptions Defense or prosecution may appeal sentences against the presumptions
Sentencing commission policy requires completion of guidelines forms State law requires departures to be accompanied by written justification Defense and pro&cut!on may appeal sentences
Sentencing commlssion policy requires completion of guidelines forms. Judges must review
State law requires departures to be accompanied by written justification Defense and prosecution may appeal sentence Information used to determine the standard range must be signed by the judge. Departures (exceptional sentences) must be accompanied by written justification Defense and prosecutlon may appeal exceptional sentences
Judges must review guidelines forms and departures must be accompanied by written
Failure to file departure may result in early release (release decision can be based on
No appellate review of guidelines sentences Judges must review sentencing orders State law requires departures to be accompanied by written justification No appellate review of guidelines sentences
Sentencing commission policy requires completion of guidelines forms Sentencing commission policy requires written justification for departures No appellate review of guidelines sentences
guidelines forms
presumptive range)
3d. Virginia
4. Missouri
Note: The table shows four basic gradations along the voluntaty/mandatory continuum. Thelower the number, the greater the extent to which a state's system is considered "mandatory".
State law requires completion of guidelines forms State law requires court clerks to forward guidelines worksheets and sentencing order to
state law requires departures to be accompanied by written justfficatlon, no appellate
Guideline forms do not have to be completed Judges may depart at their discretion
commission
review of guidelines sentences
As can be seen in Figure 1-3, at one extreme, states may use legislation
to specify all of the procedures and requirements of a sentencing guidelines
system. The sentencing commission oversees a strict system of worksheet
completion and reports departure rates or the use of aggravating and mitigating
sentences. Such a system also gives the defense and state counsel appellate
rights and uses case law to further refine the appropriate application of the
guidelines. North Carolina, for example, enacted a system that requires the
completion of a guidelines worksheet for all cases and requires sentences to fall
within a specified presumptive range. I f aggravating and mitigating
circumstances exist, judges still must sentence within specified ranges and may
not depart further from the guidelines. It could be argued that North Carolina's
system is the most "mandatory" system currently in place at the state level.
Further along the continuum, some states adopt laws that require
guidelines completion but allow departures if accompanied by written
justification. Legislation often gives the defense and prosecution the right to
appeal departures under these systems (e.g., Kansas, Minnesota, and Florida).
Virginia law requires guidelines completion and written reasons for departure,
but specifically denies appellate relief for any reason related to the guidelines.
12
Delaware's policies are similar, requiring written reasons for departure but
preventing appellate review of those sentences. Utah, rather than using state
statutes, requires completion of guidelines forms through authority of the
sentencing commission. Utah also denies appellate review for guidelines
sentences. Considered more voluntary than most, Missouri's sentencing
guidelines system does not require completion of guidelines forms and allows
judges to depart at their discretion. Michigan's guidelines are seen to fall
midway on the continuum.
Challenges to Assessing Sentencing Disparity
A central problem for the analysis of disparity is disentangling the
competing goals of sentencing from the sentencing decisions of judges.
Throughout the many characterizations of the goals of sentencing, four goals
stand out (Blumstein, Cohen, Martin, Tonry 1983, 48). First, there is a desire to
deter the offender and other potential offenders. Second, there is a desire to
incapacitate the offender so that he or she cannot commit further crimes. Third,
there is a desire for retribution against the individual for his or her social
transgressions. Finally, there is a desire to rehabilitate the offender and thereby
ensure that he or she will commit no future crimes.
Structured sentencing systems do not specify a single goal for judges to
consider when devising their sentences. Because the four goals are only partially
complementary, "the main burden of reconciling the competing goals of the
criminal justice system falls on the sentencing judge" (Hogarth 1971,4).
13
Because no two offenders or offenses are exactly alike, labeling sentencing
differences as disparity involves a moral judgment--what appears to be disparity
to one individual may simply be justifiable variation to another. Many within the
research community, however, have felt comfortable concluding that there is
evidence of disparity when similar individuals convicted of committing similar
crimes are given different sentences. A conservative approach to assessing
“similarity” is to assume that judges pay attention, in a limited fashion, to all four
goals.
Each goal requires incorporating different types of information into a study
of the sentencing decision. To pursue the goals of deterrence and/or retribution,
a judge must examine the characteristics of the offense. I f a judge wishes to
focus on incapacitation, the offender’s prior record becomes important. An
interest in rehabilitation prospects leads each judge to examine the personal
status characteristics of the offender. Researchers therefore stress the need to
include information pertinent to all of the goals…